Test Equipment Anonymous (TEA) group therapy thread

[Wolfgang]

I have seen this machine at Ham Radio in Friedrichshafen/Germany June this year. The showstopper is the lower VNA frequency of 10MHz.
Thats nothing for a radio amateur. Siglent should fix this, then it would be an interesting machine for ham radio use.

[tautech]

I have seen this machine at Ham Radio in Friedrichshafen/Germany June this year. The showstopper is the lower VNA frequency of 10MHz.
Thats nothing for a radio amateur. Siglent should fix this, then it would be an interesting machine for ham radio use.

I think it's just the limits of accuracy of the inbuilt SMD bridge. Some of the guaranteed accuracy specs are  at specific frequencies while the instrument will still operate outside those frequencies.
Still, there is zero else available in this price range that can perform SA and VNA duties.

[Housedad]

OMG.  I just caught myself.  TEA is so addictive.   (Test Equipment Acquisition)

A couple of days ago I bought the rf signal generator that has not even shipped yet  and I don't even know if it works alright yet.

I found myself looking for 2 hours or more at Spectrum analyzers. 

God help this sinner.

[mnementh]

Now it seems to are getting in a really bad mood ...
Viennese proverb, as discussed. The situation is hopeless, but no serious.

Regarding the ever-nosy Windows stuff in my Keysight instruments I decided to keep them on a USB-only island without any internet  access - that worked fine so far.
The alternative would be to constantly install windows updates, antivirus protection, all kinds of other stuff that wants to do something with you instrument while you want to measure something, ...

Another lousy invention is a instrument integration software from Keysight called BenchVue. Apart from being a huge piece of bloatware full of bugs, it looks like an early game console. No idea whom they wanted to impress with this one.

https://toolguyd.com/keysight-benchvue-is-a-great-example-of-an-unfavorable-approach-to-hardware-supporting-software/

No bad mood; just a well-needed venting. 

Much of the older os-driven stuff ran XPEmbedded. If you're lucky enough to have that (and half a gig of RAM) you might as well have an ARM/LINUX variant machine; it was pretty stable and as long as it was a slim build not optimized for internet access, ran pretty effing light & quick. Dunno about the Win7 based machines... they always struck me as like "bringing a baseball team to a chickenfight".

mnem

[mnementh]

               

Old School. Resurfaced while hunting some tools the other day. Switches & pots still feel right & firm; wiped it down but nothing more.

Who wants it before I sacrifice it to fleaBay or the Goodwill pile?

mnem
ZZZzzzZZZzzzZZZzzz...

[mnementh]

OMG.  I just caught myself.  TEA is so addictive.   (Test Equipment Acquisition)

A couple of days ago I bought the rf signal generator that has not even shipped yet  and I don't even know if it works alright yet.

I found myself looking for 2 hours or more at Spectrum analyzers. 

God help this sinner.

Abandon all hope, ye who enter here. 

mnem
Somebody had to. 

[tggzzz]

The problem is that right there... marketing wankers. You need to get rid of them, and do like Tektronix was talking about in that video on industrial design: Observe PEOPLE USING THE EQUIPMENT, and design to make it EASIER and more straightforward. THEN you will have generation after generation of die-hard fans of your product for those features.

I think it was Don Norman, now a noted usability expert, who made a useful observation about this...

To catch nmost of the problems, in most cases you only need to observe 5 inexperienced people trying to use the equipment/software/etc. If you observe those 5 people sensibly, there's not much need for large-scale and very formal studies.

[bd139]

Dunno what your fascination is with the 815 when a SVA1015X would be much better suited to your use, especially for antenna work where you can do it all with a single port connection and not need a RLB or directional coupler. That alone is a significant saving unless you have one lying around. Smith charts too ! 
Check the thread I did for my telemetry antenna build log in the RF board/forum to get some idea of just a little of what it can do.

Believe me it’s on the evaluation spreadsheet

I think it was Don Norman, now a noted usability expert, who made a useful observation about this...

To catch nmost of the problems, in most cases you only need to observe 5 inexperienced people trying to use the equipment/software/etc. If you observe those 5 people sensibly, there's not much need for large-scale and very formal studies.

Yes and no. It depends on the scope of your product and the quality of the usability panel you employ. 5 is perhaps ok for a specialised product or an internal system.

The problem you have is the inevitable bell curve. A panel of 5 doesn’t necessarily have a high enough distribution of dumbasses and people who know what they’re doing. The middle bit is mostly inconclusive as they are just zombie robots.

Average user:

https://youtu.be/H2uHBhKTSe0

[tggzzz]

I think it was Don Norman, now a noted usability expert, who made a useful observation about this...

To catch nmost of the problems, in most cases you only need to observe 5 inexperienced people trying to use the equipment/software/etc. If you observe those 5 people sensibly, there's not much need for large-scale and very formal studies.

Yes and no. It depends on the scope of your product and the quality of the usability panel you employ. 5 is perhaps ok for a specialised product or an internal system.

The problem you have is the inevitable bell curve. A panel of 5 doesn’t necessarily have a high enough distribution of dumbasses and people who know what they’re doing. The middle bit is mostly inconclusive as they are just zombie robots.

Yes and no

I think for most (not all) situations, the first 5 naive users will give the feedback on the items that require most significant changes. Beyond that it is into diminishing returns - the 80:20 effect and all that.

It does presume choosing representative users and giving them representative tasks, not just anyone doing an "unboxing and first impression". There are too many videos of the latter, and they are all ignorable as thinly disguised press releases. "Representative" should include both new customers and existing customers, inexperienced and inexperienced customers.

I wonder what would have been observed with the circular keypad? Clearly it is usable in the sense that it can be used to enter the correct values, but desirability/speed is a separate issue.

If you are after a true mass market product with knuckle dragging customers, then you have more to lose and so it is worth spending more money and time on surveys. But then I'm sure MS did that with Win8!

[Wolfgang]

... all the Keysight machinery I have (not older than 18 months max) run either Windows 7 embedded or Windows 10 IOT.
They try to make you update via internet from time to time but I refused. So far - no effect, they just work.

[mnementh]

I think it was Don Norman, now a noted usability expert, who made a useful observation about this...

To catch nmost of the problems, in most cases you only need to observe 5 inexperienced people trying to use the equipment/software/etc. If you observe those 5 people sensibly, there's not much need for large-scale and very formal studies.

Yes and no. It depends on the scope of your product and the quality of the usability panel you employ. 5 is perhaps ok for a specialised product or an internal system.

The problem you have is the inevitable bell curve. A panel of 5 doesn’t necessarily have a high enough distribution of dumbasses and people who know what they’re doing. The middle bit is mostly inconclusive as they are just zombie robots.

Yes and no

I think for most (not all) situations, the first 5 naive users will give the feedback on the items that require most significant changes. Beyond that it is into diminishing returns - the 80:20 effect and all that.

It does presume choosing representative users and giving them representative tasks, not just anyone doing an "unboxing and first impression". There are too many videos of the latter, and they are all ignorable as thinly disguised press releases. "Representative" should include both new customers and existing customers, inexperienced and inexperienced customers.

I wonder what would have been observed with the circular keypad? Clearly it is usable in the sense that it can be used to enter the correct values, but desirability/speed is a separate issue.

If you are after a true mass market product with knuckle dragging customers, then you have more to lose and so it is worth spending more money and time on surveys. But then I'm sure MS did that with Win8!

The problem is this: That 80% is the difference between a product that mostly works and a product that is a joy in the hand. The difference between a Yugo and a Rav4, and again between the Rav4 and a Testarossa.

There's "It works. Sortof. Ship it." Then there's "It works right. We're ready for the big release." Then there's "OMG, this thing is almost PERFECT; it's the best thing since electric lights and running water! I can't WAIT to bring this to the people!"

Used to be we strove for "OMG!", we expected and usually got "It works right." and occasionally ran short and had to settle for "It works. Sortof."

Now we feel lucky if we get "It works. Sortof.", "It works right." is just for silly perfectionists and "OMG!" only EVER happens by accident.

There's an episode of South Park called "Raising the Bar"; it is highly critical, even downright mean towards obesity as it is happening in America. However... it is spot-on in regards the general nosedive that everything has taken in terms of what is considered "minimum acceptable".

The main reason is the shift from a manufacturing economy to a consumer economy... but that's a reason, not an EXCUSE.

If you've never made ANYTHING, how can you have pride in your work? How can you even know what a minimum standard IS?

mnem
When did basic competency at one's job become "a goal"? When did a dumb blank stare become an acceptable defense for global incompetency?

[bd139]

Exactly. I see the 80/20 rule quoted and adhered to in lots of places but I’ve seen a lot of customers lost over the niggling 20% of problems remaining. The analogy I like to use, particularly when pitching my services is every 5th step you take lands in a pile of dog shit.

I see the term deminishing returns used also. But people forget that customer retention,  satisfaction and referrals are quantifiable returns.

Last project I was on I hand held a “minor unimportant customer” as I was told they were through a bunch of problems they had. The guy quit there about a month later and started at a big corp immediately and pitched the product to them. £17m for that 20%. Not bad.

Product is done when it’s done not when it’s 80% done. And just because some of it is software doesn’t mean you can update it later. That’s a shit excuse.

[mnementh]

 

Whenever I hear you talk about software, I'm reminded of this XKCD now. Especially the part about blockchain. 

mnem
Software only fixes problems with wetware; it doesn't fix fucked-up hardware. ~me

[tggzzz]

I think it was Don Norman, now a noted usability expert, who made a useful observation about this...

To catch nmost of the problems, in most cases you only need to observe 5 inexperienced people trying to use the equipment/software/etc. If you observe those 5 people sensibly, there's not much need for large-scale and very formal studies.

Yes and no. It depends on the scope of your product and the quality of the usability panel you employ. 5 is perhaps ok for a specialised product or an internal system.

The problem you have is the inevitable bell curve. A panel of 5 doesn’t necessarily have a high enough distribution of dumbasses and people who know what they’re doing. The middle bit is mostly inconclusive as they are just zombie robots.

Yes and no

I think for most (not all) situations, the first 5 naive users will give the feedback on the items that require most significant changes. Beyond that it is into diminishing returns - the 80:20 effect and all that.

It does presume choosing representative users and giving them representative tasks, not just anyone doing an "unboxing and first impression". There are too many videos of the latter, and they are all ignorable as thinly disguised press releases. "Representative" should include both new customers and existing customers, inexperienced and inexperienced customers.

I wonder what would have been observed with the circular keypad? Clearly it is usable in the sense that it can be used to enter the correct values, but desirability/speed is a separate issue.

If you are after a true mass market product with knuckle dragging customers, then you have more to lose and so it is worth spending more money and time on surveys. But then I'm sure MS did that with Win8!

The problem is this: That 80% is the difference between a product that mostly works and a product that is a joy in the hand. The difference between a Yugo and a Rav4, and again between the Rav4 and a Testarossa.

There's "It works. Sortof. Ship it." Then there's "It works right. We're ready for the big release." Then there's "OMG, this thing is almost PERFECT; it's the best thing since electric lights and running water! I can't WAIT to bring this to the people!"

Used to be we strove for "OMG!", we expected and usually got "It works right." and occasionally ran short and had to settle for "It works. Sortof."

Now we feel lucky if we get "It works. Sortof.", "It works right." is just for silly perfectionists and "OMG!" only EVER happens by accident.

There's an episode of South Park called "Raising the Bar"; it is highly critical, even downright mean towards obesity as it is happening in America. However... it is spot-on in regards the general nosedive that everything has taken in terms of what is considered "minimum acceptable".

The main reason is the shift from a manufacturing economy to a consumer economy... if you've never made anything, how can you have pride in your work? How can you even know what a minimum standard IS?

mnem
When did basic competency at one's job become "a goal"? When did a dumb blank stare become an acceptable defense for global incompetency?

I think that for the example stated (circular keypad) and others, the 80% would include things like "why are you doing it that way?". Provided the test panel isn't only fresh-out-of-school never-used-anythig-like-it-before weenies.

But don't believe amateur UI designers like me and others (and you!). Do believe someone with a solid reputation in assessing human-machine interaction, e.g. Don Norman.

[mnementh]

No longer a professional is not the same as amateur.

mnem
Professional asshole. Retired.

[tggzzz]

Exactly. I see the 80/20 rule quoted and adhered to in lots of places but I’ve seen a lot of customers lost over the niggling 20% of problems remaining. The analogy I like to use, particularly when pitching my services is every 5th step you take lands in a pile of dog shit.

Most soundbite analogies are like misused statistics - designed to conceal rather than illuminate.

I see the term deminishing returns used also. But people forget that customer retention,  satisfaction and referrals are quantifiable returns.

Last project I was on I hand held a “minor unimportant customer” as I was told they were through a bunch of problems they had. The guy quit there about a month later and started at a big corp immediately and pitched the product to them. £17m for that 20%. Not bad.

Product is done when it’s done not when it’s 80% done. And just because some of it is software doesn’t mean you can update it later. That’s a shit excuse.

Well, all that is true, but it conceals as much as it illuminates.

If you like soundbites then "works of art are never "done", merely abandoned". The issue is the definition of "done"; that has to be pragmatic. Yes, people do get it wrong sometimes; is anybody surprised?

[bd139]

I do the illumination after I've pimped myself out

You're right on the pragmatism of done. Is it mature, is it minimum viable product, is it zero bug bounce. That depends on what you are working on.

[mnementh]

I think that for the example stated (circular keypad) and others, the 80% would include things like "why are you doing it that way?". Provided the test panel isn't only fresh-out-of-school never-used-anythig-like-it-before weenies.

But don't believe amateur UI designers like me and others (and you!). Do believe someone with a solid reputation in assessing human-machine interaction, e.g. Don Norman.

But CLEARLY, that question was NEVER asked. If the 80/20 rule WORKED as you suggest things like this would be an aberration, NOT THE RULE as they obviously are.

The problem with that 80/20  BS is that it shifts the bar LOWER EVERY TIME IT IS APPLIED; and it gets applied recursively ad infinitum, because the decisions are made by people (MBAs) whose entire "education" consists of learning the current buzzwords and making numbers lie, NOT by the people who have to use something and work with it everyday.

Eventually "Does it work?" gets shifted to the wrong side of "diminishing returns". INEVITABLY.

With all due respect to Don, you can find experts who believe and will say any damned thing you like. He's WRONG about usability studies, and the only proof you need of that is to look at the things we all have to use everyday. Even toasters are a useless mess.

Cheers,

mnem
"...I am an expert witness, because I say I am. And I said,
'Gentlemen... (and I use that word loosely)
I will testify for you. I'm a gun for hire, I'm a saint, I'm a liar;
Because there are no facts, there is no truth, just data to be manipulated.
I can get you any result you like... what's it worth tooya?'
Because there is no wrong, there is no right, and I sleep very well at night..." ~ Don Henley - In the Garden of Allah

[bd139]

Yes. Don't start me on fucking toasters. My toaster is one of these: https://www.amazon.co.uk/dp/B016FU6WLY/

To quote an accurate review:

"Basically an OK toaster, but annoying features like: you can't use the second slots if the first slots aren't being used, also if you put two bits of bread in slots 1 and then a couple of minutes later put some in slots 2, the second lot will pop up when the first are done. Stupid feature."

They broke the usability pattern established with the toaster for appearance, cost and/or innovation purposes. 

I think the seller was feeling slightly guilty about my 87V and posted it special delivery next day before 1PM. Should get it Monday. I don't blame them. We all have bad weeks

Edit: relatives have finally buggered off (woohoo), the kids are amusing themselves now after massive sugar crash so I'm going to go and play with RF amps and analyse them in the frequency domain. It's wierd but until you get an SA you don't realise how blind you are to the workings of some circuits. Revisiting older things I built from now to fix my brain damage from time domain only measurements. Amazingly useful bit of kit.

[Wolfgang]

Hi,

the list is imcomplete.

1. Buy essential test equipment at bargain price
2. Investigate faults.
3. Fix major faults
4. Buy a similar model for parts
5. Repair both, and keep both in case one breaks
6. Find minor faults in functions you'll never use
7. Buy another piece of test equipment to fix / calibrate it with
8. goto 2

Between 1 and 2 there should be a "Buy an industry standard new big iron so you have a reference what is right or wrong." Then you can skip 4 to 7. In the time saved; Go for even more TEA stuff ! 

[bd139]

I'm at step 4. Looking for Tek 7000 spares now 

The guy who sold me my 7603 at the hamfest for £50 had another 7603 for £50 and a storage 7k mainframe for £50. Starting to think I should have lifted another one

[Mr. Scram]

... all the Keysight machinery I have (not older than 18 months max) run either Windows 7 embedded or Windows 10 IOT.
They try to make you update via internet from time to time but I refused. So far - no effect, they just work.

Interesting. I thought Keysight was into CE and was preparing to possibly drop that.

[mnementh]

Hi,

the list is imcomplete.

1. Buy essential test equipment at bargain price
2. Investigate faults.
3. Fix major faults
4. Buy a similar model for parts
5. Repair both, and keep both in case one breaks
6. Find minor faults in functions you'll never use
7. Buy another piece of test equipment to fix / calibrate it with
8. goto 2

Between 1 and 2 there should be a "Buy an industry standard new big iron so you have a reference what is right or wrong." Then you can skip 4 to 7. In the time saved; Go for even more TEA stuff ! 

That only applies if you've never used stuff that works right.

Most of the folks here have been around long enough and have worked corporate at some point and so had access to "The good stuff", and WE KNOW THE DIFFERENCE. That is the core of the discussion on "minimum standards" we're having right now.

mnem
I'm currently considering step 4, only with respect to Franken-Cruiser. 

[mnementh]

I'm at step 4. Looking for Tek 7000 spares now 

The guy who sold me my 7603 at the hamfest for £50 had another 7603 for £50 and a storage 7k mainframe for £50. Starting to think I should have lifted another one

Yeah, sorry; you fucked up on that one. 

To quote the old Pentagon procurement adage: "Why buy only one of something, when you can have two for twice as much? And a spinoff project for three times that?"

mnem
Mmm... parts salad. Yummy.

[bd139]

... all the Keysight machinery I have (not older than 18 months max) run either Windows 7 embedded or Windows 10 IOT.
They try to make you update via internet from time to time but I refused. So far - no effect, they just work.

Interesting. I thought Keysight was into CE and was preparing to possibly drop that.

CE no longer exists. No new licenses being granted. Phasing out support in 2023. That's the only reason why. On embedded it wasn't a bad product. Better than NT which is a bloated turd.

I'm at step 4. Looking for Tek 7000 spares now 

The guy who sold me my 7603 at the hamfest for £50 had another 7603 for £50 and a storage 7k mainframe for £50. Starting to think I should have lifted another one

Yeah, sorry; you fucked up on that one. 

To quote the old Pentagon procurement adage: "Why buy only one of something, when you can have two for twice as much? And a spinoff project for three times that?"

mnem
Mmm... parts salad. Yummy.

Yep tell me about it. Had room in the car and the cash on me as well